FX-Sabers.com
The S.A.B.E.R. GUILD: Saber Manufacturers => Plecter Labs => Sound boards technical issues / questions / problems => Topic started by: AngelOmega on January 27, 2020, 11:47:37 AM
-
I recently bought a CFX through the Custom Saber Shop. I’d been planning a Crystal Focus build for a few years now, so I’m familiar with the proper wiring, and adjusting to the new board did not require much adaptation of my precious plans.
However, when I first booted up the board an turned on the power, I got an ignition sound, followed by a brief 1 second hum, before my saber began to lose its mind. Swings and clashes kept repeating at startling speed even while the saber was sitting still.
Following advice from another forum, I made sure all my connections were good, I made sure the battery was fully charged, and I hooked it up to RICE to see what the problem might be. The RICE program told me that the board was receiving near constant swing inputs.
At this point it was very late at night, and I resolved to try a firmware update in the morning. The next morning however, upon turning the saber on, I no longer got the constant swing sounds. Instead, the saber seemed to function fine, and the clashes worked... but I didn’t get any swing sounds. It would only trigger impact sounds.
I went ahead with a firmware update, which failed to solve the issue. Quite the opposite. After the firmware update, the LEDs have quit working. I’ve tested the LEDs alone with a battery to make sure they were still functional, and they light up just fine. The board seems to have stopped sending signal to the LEDs.
I believe I’ve exhausted my ability to troubleshoot the board, though it pains me to give up. I don’t want to continue trying to work on it without knowing exactly what is wrong. I think it must be a faulty swing sensor, but I’m unsure how to go about repairing it.
I’ve talked with others on the TCSS forum about similar issues they’ve had, though a solution has yet to be found. I’m not quite sure where to go from here.
Help me Forum Members, you’re my only hope.
-
For your leds, strippower was introduced, and needs to be reverted to 0000 for hb leds (see changelog).
For the swings, you might want to check the board voltage map. One possible failure is the capacitor located with the 14v on the voltage map.
You can see if replacing it solves the issue.
Otherwise you can talk with Chewbacca here for diags and repairs
-
Thanks for the quick response! Solved my led issue! I checked the backup file of the sd that I have on my computer and that was in fact the issue.
As for the capacitor, do you by any chance have the specs for the capacitor should I decide to try to replace it? I’m away from the board for the next day or so, otherwise I’d identify the proper one myself.
Thanks so much for the help!
Edit: I’m unsure what you mean by voltage map. Is that something in RICE? If so, how do I access it?
-
my bad for not putting the specs yesterday (on my phone, about to go to bed). You'll find the voltage map in the manual. The capacitor is the one next to the 14V test point, it's a 1uF / 25V capacitor, ceramic X5R (or X7R) in 0603 footprint.
Mouser reference provided as an example : 963-TMK107BLD105KA-T
We got a few capacitors that ended up with internal fracture for unknown reasons. That's the first thing to check and replace before considering looking at the sensor itself, which could be dead by ESD.
-
My soldering station is ESD safe, so likely its the capacitor. Thanks for the help!
-
I’ve replaced the capacitor, and still no sound. I assume this means the swing sensor must be damaged. I contacted Chewbacca about repairs and am waiting on a response.
-
ok, keep us posted. that's really a one of a kind, as you had swings working.
-
I am currently having almost the exact same problem, except that I'm using a neopixel blade rather than the in-hilt LEDs. The saber was working fine, but a couple times it started giving me the constant swing sounds. I was able to resolve that issue simply by power cycling the board. However, after leaving the saber for a night, by the morning I was unable to get any swing sounds to play at all. The board still recognizes clashes and board orientation (as far as I can tell), but not swings.
Should mention I also got my board through TCSS. It was meant to be a replacement board after I accidentally fried the audio amplifier on my previous one. I'm really hoping to find a solution to this without having to replace the board yet again, but it definitely seems like this is a hardware issue and not solvable through software.
-
I am apparently In the same boat. Everything is good, including clashes, but zero swings on smoothswing or legacy. I took Erv’s latest FW that logs the gyro output, and the gyro output goes bonkers on boot and then dead on saber ignition.
Erv recommended the capacitor replacement. So I ordered that today and fingers crossed that it fixes the issue. Really bummed to be dead in the water on this one.
-
I am currently having almost the exact same problem, except that I'm using a neopixel blade rather than the in-hilt LEDs. The saber was working fine, but a couple times it started giving me the constant swing sounds. I was able to resolve that issue simply by power cycling the board. However, after leaving the saber for a night, by the morning I was unable to get any swing sounds to play at all. The board still recognizes clashes and board orientation (as far as I can tell), but not swings.
Should mention I also got my board through TCSS. It was meant to be a replacement board after I accidentally fried the audio amplifier on my previous one. I'm really hoping to find a solution to this without having to replace the board yet again, but it definitely seems like this is a hardware issue and not solvable through software.
We confirmed it's not software issue and haven't been able to reproduce the capacitor break in the lab but after changing it, it did work again on the 2 boards I was sent.
-
Thanks for the update Erv, I’m glad to hear replacing the capacitor could be enough to fix the issue. I can give that a shot and see how it goes.
Just for my peace of mind, I want to confirm that I’m understanding the correct part to replace. I’m linking an image where I’ve circled the part I believe needs to be replaced. Could you potentially confirm that I’m correct before I go ahead with the swap?
(https://i.ibb.co/WNhFRdT/6-E819619-A4-F1-4-EDB-BC96-8-DE3519-C1668.jpg) (https://ibb.co/WNhFRdT)
Thank you!
-
yes, that's the one !
-
Great, thank you!
-
Replacing the Capacitor didn’t seem to solve the issue for me, however I admit I had a hard time replacing the capacitor (not used to soldering things that small). It is possible the connections I made weren’t adequate. Either way, it has been sent off to Chewbacca for diagnosis and repairs.
-
We confirmed it's not software issue and haven't been able to reproduce the capacitor break in the lab but after changing it, it did work again on the 2 boards I was sent.
I’ve just finished swapping out the capacitor on my board, and the motion problem has persisted. I used the spare capacitor from my previous board with the audio amp issue, but that board had no issues with motion sensing so I’m assuming the capacitor part was good.
I can always try to order a brand new one to try as well, but I figured I’d post an update at this point.
It’s interesting that the two boards you were sent were fixed via a new capacitor, but the now two people in this thread so far that have tried weren’t able to fix the problem. I wonder why that is.
Thanks for all the help so far! I really hope we can get to the bottom of this issue.
-
I had 2 to repair and the gyro was dead from ESD from what I could see. I first removed the sensor to re-tin the pads and re assembled, I got it to work in the sense the sensor was found by the MCU (identified with the 2 tokens during init/boot), mag sensor worked for magic ignition, accelerometer was working, gyro silent. I replaced the sensor with a new one in those 2 cases.
-
I wonder if the initial cause of the issue was a bad capacitor, and perhaps that led to a damaged swing sensor? In other words, the ones that were fixed by a replaced capacitor had intact swing sensors, yet our swing sensors were damaged due to bad current through the broken capacitor. I know I tried turning off/on my saber, rebooting it, reformatting the sd, etc, many times before I replaced the capacitor. It may be that continued usage with the bad capacitor messed up the swing sensor after some time.
-
We suspect ESD, in particular from the USB port, which also explains the situation for which the user updates his saber (usually with no/little prior use of the USB or RICE, using SD card editing instead or just not touching anything) and get the problem on the gyro. We have ESD (some) protection on the USB but might need to make it even more bulletproof in the future. My computer uses a docking station with a USB hub, proper house Mains and earthing and those things didn't happen to me (even on prior board gen with fried act or aux pads).
ESD via the computer / USB is common, but seems worse in certain situations / installations
-
That would definitely make sense for me actually. I wasn’t really aware of concerns in using RICE, but connected my saber to my computer many times to check and edit the board. Something must have happened one (or several) of the times that I hooked it up.
I’m terms of replacing the motion sensor, do you think that’s something I’d reasonably be able to handle myself with just a regular soldering iron? Or does it require more specialized equipment and expertise? I still have the one from my other board, which should still work in theory.
-
I'm actually glad I have been patrolling the forums while I wait for my CFX saber parts. I plan on making a remote connection in hilt for the CFX USB, so now I know I'll need to take ESD into account when making the connector.
-
I’m terms of replacing the motion sensor, do you think that’s something I’d reasonably be able to handle myself with just a regular soldering iron? Or does it require more specialized equipment and expertise? I still have the one from my other board, which should still work in theory.
doable with a hot air station and liquid (rework) flux. *NOT* easy task, I sometimes need multiple rework until I can bring the PCB at the right temp. before it all self centers on the pads.
You can't replace the motion sensor with just a soldering iron.
-
doable with a hot air station and liquid (rework) flux. *NOT* easy task, I sometimes need multiple rework until I can bring the PCB at the right temp. before it all self centers on the pads.
You can't replace the motion sensor with just a soldering iron.
Understood, thanks for the advice. It’s definitely not something I’ll be able to handle myself then, as I am still quite an amateur.
In that case, what are the steps I can take to get the board repaired? I saw some people mention “Chewbacca” but how can I contact them? And what are the associated costs?
Thank you so much for all the help!
-
this is michael's handle here on the forum, just PM his with that character name. Usually a repair is in the range of $20-30 shipping included
-
Erv -
Add me to the list of capacitor replacements that did not solve the no swing issue. I received the capacitor and installed it. Not pretty but think it has solid connections. Clash still works (assuming that means the cap was attached successfully) but no legacy or smoothswing.
-
ok scott thank you. It must then be a cold joint under the motion sensor :sad: This should be covered by warranty, you can PM chewbacca here to get it fixed. Does the gyro logs changing values ? (latest fw + latest RICE + remote tab) or it is rather static ?
Last question : have you tried pressing a little on the sensor itself while the boards is powered and logging motion in RICE ?
-
Definitely didn’t try pushing on the motion sensor, but I will give that a shot this evening. I also did not use the latest FW, but the version with the “gyro” suffix that you provided. Still static. If it is if any value, I can flash the latest package and look at it with the RICE editor that you added the line graph monitor to.
-
yes, give a try to the latest fw with the motion log, and try pressing, that will help to diag, thank you for your time on this
-
No worries, will do this this evening and capture any data/feedback that I can, glad to be able to help the root cause analysis in any way I am able.
-
something interesting, I got one board to work again temporarily then crazy gyro again. I have corrected the manual for the voltage map, I confirm that the proper voltage for the gyro cap is rather 9-10v (not 14-15v).
I discovered that the motion sensor works again with NO capacitor which could indicate some sensors internal charge pump have an issue with the capacitor value. As a result, the gyro high voltage isn't buffered which isn't ideal but the board works. That's sort of a relief as it doesn't indicate any manufacturing issue (soldering / cold joint of the sensor). I don't know why yet it behaves like this, my dev board and 2 other sabers have a stock board and components (including that capacitor) with no trouble for months, so it seems to be sporadic.
Would you guys with that issue mind trying with no gyro cap ? thank you
-
I just sent over the new logs, but will pull the replacement cap back off and report back!
-
I just removed the cap and the same behavior persists. Clash is good, no swing sounds.
-
Finally gave up...swapped the board out of my build with the same wiring and same config (albeit, I left the FW alone given the posts about USB port ESD) and all is perfect.
Now just need to get this wonky board fixed.
-
I just attempted removing the capacitor as well, and no luck. Still no swings. I haven’t been able to check the gyro log as I don’t think that version of the RICE editor is available for Mac yet.
Haven’t heard back from Chewbacca yet on the forum, so I’ll probably drop him an email to inquire about repairs.
-
Yes I had to contact Chewbacca directly through email. He responded very quickly once I did.
-
I have a new CFX neopixel install that also has the same behavior as described by others. No swing sounds or smoothswing but normal hum and clash. I upgraded the FW to 10.004E with default fonts and config files for 10.004D hoping it was software related. The truncated boot log from RICE shows:
SPECS:Sectors 15523840 (7680MB)
SD Clock (CSD) 25000000 Hz
SD Clock 50000000 Hz
Type SD
SD Version 3.
High Speed
Card Installed
Initialized
Plecter Labs Crystal Focus
v10.004E-05/02/2020
Init Motion Sensor
Reset Motion Sensor
XG boot ended: OK (waited 1ms)
M boot ended: OK (waited 1ms)
WHO_AM_I_M = 3D - WHO_AM_I_XG = 68
Success
Motion sensor Debug:
CTRL_REG1_G = 10111000
CTRL_REG2_G = 0
CTRL_REG2_G (second time) = 0
CTRL_REG3_G = 0
CTRL_REG4 = 111000
I also measure 8.71V at the capacitor with my DMM which looks to be in the expected range.
Reading other posts, people have:
1.Looked at gyro info with custom FW from erv to see if gyro displays appropriate motion.
2.Replaced the 25V 1uF 0603 capacitor with SMB footprint or remove it entirely.
3.Sent it in for gyro replacement.
Would it be helpful to replicate what others have done in the interest of shared knowledge? I want to buy another sound modules for an install but don't want to have the same issues if there is something I can do to prevent it like extra ESD equipment I should purchase.
-
after digging several situation and testimonials from customer, it's probably ESD that damages the gyro charge pump leading to either silent or crazy gyro. You can try to remove the gyro cap (pointed where it's currently written 14V, where 8-10v should rather be seen).
My conclusion is that people discharge ESD in the capacitor or in the switches when manipulating (killing the act or aux pad internal pullup resistor).
The fw has been ruled out as it writes properly in the sensor registers and init goes perfect, even reading back from the sensor work, just the internal MEMS is damaged.
-
So I’ve heard back from Chewbacca, and the issue was in fact caused by a damaged motion sensor. I think it is safe to call this issue solved now, as it seems we have all reached the same conclusion that the motion sensor was damaged by ESD.
-
Got my board back from Chewbacca. Soldered it up and got it all connected and packed into the saber.
Everything works great! No issues. Only thing left is the process of dialing in the sensitivities and fx timings the way I like them!
Love the board! This is my first saber with sound!
-
How did you get in touch with Chewbacca?
-
How did you get in touch with Chewbacca?
He’s a moderator here, shoot him a PM.
-
Got my board back from Chewbacca. Soldered it up and got it all connected and packed into the saber.
Everything works great! No issues. Only thing left is the process of dialing in the sensitivities and fx timings the way I like them!
Love the board! This is my first saber with sound!
good news thanks for the notification !(and enjoy your saber project)
-
Thanks K2S0....that was my original approach, just tried again.
-
He replied to direct email very quickly. His email is linked under his profile.